A recent CDC report observed that heroin use has increased in men and women across the U.S., including most age groups and all income levels. The current heroin epidemic follows an ongoing prescription opioid misuse epidemic that has helped drive the spike in heroin use. Research to improve understanding of the pathophysiology of opioid use disorder (OUD) is thus a public health imperative. OUD is a moderately heritable (40-60%) disorder; however, few replicated genetic findings have been identified. One reason for this is that a definitive control sample for association studies of OUD has not been established. Building on candidate gene data findings from the Comorbidity and Trauma Study (CATS) that noted association results for OUD vary substantially with substance exposure in the comparison group, and consistent with well-replicated findings from comparisons of non-dependent, regular smokers to nicotine dependent individuals, we conducted a genome-wide association study (GWAS) in CATS limiting controls to opioid misusers. The strongest association signal observed was for cornichon family AMPA receptor auxiliary protein 3 (CNIH3) SNPs. Replication performed in two additional GWAS samples yielded a genome-wide significant (GWS) meta- analytic p value of 4.3 E-9 for rs10799590, the most strongly associated SNP. These findings, coupled with exciting results from a pilot meta-analysis included in this revised application demonstrate the utility of comparisons to samples of non-OUD opioid misusers (even those with modest N's) to identify novel loci that would not be detected in comparisons to general population controls. The current application proposes to ascertain, assess, and genotype groups [African American (AA) N=1000 and European American (EA) N=1000] of individuals with a limited history of opioid misuse who have never met criteria for an opioid use disorder (i.e., non-OUD misusers as controls). Individuals with OUD (N=400 AA and N=400 EA) will also be assessed, and genotyped and combined with OUD cases from other sources for a GWAS. The specific aims are: Aim 1) To ascertain, comprehensively assess, and obtain DNA from 1000 AA and 1000 EA individuals with a history of limited opioid misuse who have never met criteria for opioid use disorder (OUD) and smaller numbers of individuals with OUD (400 AA and 400 EA) OUD - the first effort targeting large-scale collection of non-OUD opioid misusers to serve as controls. Aim 2) To conduct mega-analyses: 1) comparing the current proposal's AA and EA non-OUD opioid misusers to expanded groups of OUD cases that also include previously GWAS-genotyped samples, AA (N=2093) and EA (N=6154); 2) with further expanded samples of both cases and controls (AA totals: 4366 OUD cases; 1699 non-OUD misuser controls; EA totals: 10,106 OUD cases; 2138 non-OUD misuser controls) from other sources. Aim 3) To examine whether subjective responses to initial opioid use and stages of use mediate or moderate the relationship between top SNPs from the mega-analysis and progression to OUD.